Gone With the Wind: Consumer Surplus from Renewable Generation

Transcription

1 Gone With the Wind: Consumer Surplus from Renewable Generation Matt Butner Department of Economics University of Colorado Boulder USAEE North American Conference September 24, 2018

2 Increase in Renewable Electricity Generation Capacity Implications: - Decreased pollution - Increased intermittent electricity generation - Increased low marginal cost generation

3 Increase in Renewable Electricity Generation Capacity Implications: - Decreased pollution - Increased intermittent electricity generation - Increased low marginal cost generation

4 Wind generation impact on the Merit Order Merit Order Effect of Renewable Generation $/MWh Demand Oil Supply P 0 Coal Gas Nuclear Q MWh

5 Wind generation impact on the Merit Order Merit Order Effect of Renewable Generation $/MWh Demand Oil Supply P 0 P 1 Coal Gas Nuclear Wind Q MWh

6 This Decrease in Price is Good for Consumers According to a May 2012, MISO-Commissioned study by Synapse Energy Economics, Inc., in addition to the benefits to local economics, wind has show to reduce overall energy costs for consumers saving ratepayers $63 to $147 per year (assuming a 20 GW scenario in 2020)...

7 This Decrease in Price is Good for Consumers According to a May 2012, MISO-Commissioned study by Synapse Energy Economics, Inc., in addition to the benefits to local economics, wind has show to reduce overall energy costs for consumers saving ratepayers $63 to $147 per year (assuming a 20 GW scenario in 2020)... - However, this assumes perfectly competitive markets.

8 Research Questions 1. How much do consumers benefit from reduced operating cost associated with increased renewable generation? - This will depend on how firms change their competitive strategies in response to increased renewable generation

9 Research Questions 1. How much do consumers benefit from reduced operating cost associated with increased renewable generation? - This will depend on how firms change their competitive strategies in response to increased renewable generation 2. How do firm s change their strategy in response to increased renewable generation? - There is an incentive for firms to withhold their output when their wind turbines are generating electricity.

10 Research Questions 1. How much do consumers benefit from reduced operating cost associated with increased renewable generation? - This will depend on how firms change their competitive strategies in response to increased renewable generation 2. How do firm s change their strategy in response to increased renewable generation? - There is an incentive for firms to withhold their output when their wind turbines are generating electricity. Results: 1. Consumer benefit can be up to 68 USD per person per year 2. Observed withholding reduces consumer surplus by > 30%.

11 Theory: How Wind Generation Impacts the Price Market equilibrium: d(p) = }{{} Demand S o (p) o }{{} Conventional Supply + W }{{} Wind Differentiating the market equilibrium with respect to wind generation provides: dp dw = 1 + S o(p) o W S o (p) d (p)

12 Theory: How Wind Generation Impacts the Price Market equilibrium: d(p) = }{{} Demand S o (p) o }{{} Conventional Supply + W }{{} Wind Differentiating the market equilibrium with respect to wind generation provides: dp dw = 1 + S o(p) o W S o (p) d (p) - I observe S o(p) and d (p) - I need firm theory to find So(p) W

13 Firm s incentives to withhold output $/MWh Demand Supply P 0 Q MWh Trade off between P and S o (p) or P and S o (p)

14 Firm s incentives to withhold output $/MWh Demand Supply P 1 P 0 Q MWh Trade off between P and S o (p) or P and S o (p)

15 Diverse firm s incentives to withhold output $/MWh Demand Supply P 0 Q MWh Benefits increase when you own wind generation, math proof

16 Diverse firm s incentives to withhold output $/MWh Demand Supply P 1 P 0 Q MWh Benefits increase when you own wind generation, math proof

17 Theoretical predictions We expect 1. Only diverse market participants, owning wind turbines and other assets, will withhold their generation offer from their traditional units in response to wind generation. 2. Market participants that own more wind generation capacity will withhold their traditional units more. 3. Market participants will withhold more in response to wind generation from their own wind turbines.

18 MISO Wholesale Electricity Market, Multi-unit uniform price auction - 70 GWh on average - 1/2 coal, 1/4 gas - 5 to 15 GWh of wind - Average 27 $/MWh - Locational Marginal Price - Energy - Losses - Congestion Supplemental information: Owner Portfolios Wind PPAs Vertical Arrangements

19 Testing for Physical Withholding, Estimating S o(p) W Use hourly ex-ante supply curves of all market participants - Aggregate supply curves at the owner level - Exclude bids from wind turbines - Interpolate / Extrapolate supply curve on a common domain - Data are q otb, p b Estimate δ in the following equation q otb =δwindgwh t + X β + η opb ymh + ε otb - X includes load, net exports, congestion, wind forecast error, natural gas prices, temperature. sources. - η opb ymh is owner, year-month-hour, average supply curve If δ < 0 then the conventional assets are being withheld.

20 Variation in Supply Offer Curves All supply curves, by owner o, hour h, month m, year y.

21 Result 1 - Withholding Full Sample Firm s that own wind turbines withhold in response to more W. Wind GWh, δ Doesn t Own Wind Wind GWh Owns Wind Wind GWh Quantity Offered, MWh (0.558) (0.163) (3.728) Owner-Price-Year-Month-Hour Fixed Effects Yes Yes Observations 28,811,160 28,811,160 R-squared Source: MISO Real Time Offer Market Reports January 1, 2014 to December 24, Peak hours, defined as 3pm to 8pm inclusive. Offer curves are interpolated and defined at $3 intervals between 0 and 60 USD. Standard errors, in parenthesis, are clustered by month of sample and owner.,, denote p-value less than 0.1, 0.05, and 0.01 respectively.

22 Result 2 - Market Participant Specific ˆδ Firm s that own more wind generation, withhold more.

23 Result 3 Firm s that own wind withhold more in response to their own wind

24 Implications for Consumer Surplus Consumer surplus from electricity, hour t at market price p CS t (p) = p D t (x)dx implies the total change in consumer surplus is CS = t D t (p) dp dw t. (1) dw t I directly calculate two versions of dp dw t - No withholding, perfect competition - Observed withholding, using estimates ˆδ o Details on dp dw Statistics of dp dw Reconstructing Equilibrium

25 Consumer Surplus Calculation, MISO, Consumer Surplus Total, Billion $ Annual $/person Expenditure CS comp, no withhold CS obs, obs. withhold CS comp CS obs Notes: Time period of interest is from January 1st, 2014 to December 24th, Revenue is the sum of Market MEC and market generation quantity in MWh for all hours. The equilibrium is where supply net of wind equals total demand within MISO. Annual per person calculations divides the total quantity by 2.98 years and 50 million people. All numbers are in nominal US dollars.

26 Consumer Surplus Calculation, MISO, Consumer Surplus Total, Billion $ Annual $/person Expenditure CS comp, no withhold CS obs, obs. withhold CS comp CS obs Notes: Time period of interest is from January 1st, 2014 to December 24th, Revenue is the sum of Market MEC and market generation quantity in MWh for all hours. The equilibrium is where supply net of wind equals total demand within MISO. Annual per person calculations divides the total quantity by 2.98 years and 50 million people. All numbers are in nominal US dollars.

27 Consumer Surplus Calculation, MISO, Consumer Surplus Total, Billion $ Annual $/person Expenditure CS comp, no withhold CS obs, obs. withhold CS comp CS obs Notes: Time period of interest is from January 1st, 2014 to December 24th, Revenue is the sum of Market MEC and market generation quantity in MWh for all hours. The equilibrium is where supply net of wind equals total demand within MISO. Annual per person calculations divides the total quantity by 2.98 years and 50 million people. All numbers are in nominal US dollars.

28 Consumer Surplus Calculation, MISO, Consumer Surplus Total, Billion $ Annual $/person Expenditure CS comp, no withhold CS obs, obs. withhold CS comp CS obs Notes: Time period of interest is from January 1st, 2014 to December 24th, Revenue is the sum of Market MEC and market generation quantity in MWh for all hours. The equilibrium is where supply net of wind equals total demand within MISO. Annual per person calculations divides the total quantity by 2.98 years and 50 million people. All numbers are in nominal US dollars.

29 Policy Implications According to a May 2012, MISO-Commissioned study by Synapse Energy Economics, Inc., in addition to the benefits to local economics, wind has show to reduce overall energy costs for consumers saving ratepayers $63 to $147 per year (assuming a 20 GW scenario in 2020)... I find a potential consumer benefit of $68 per year, with 17 GW of capacity. However, $22 per person per year is lost to uncompetitive behavior by electricity generators.

30 Discussion - It s important to have competitive markets - That s the goal of the Federal Energy Regulatory Commission and Independent System Operators - This is due to how markets are structured - Alternative pricing agreements will reduce the incentive to withhold. - However, should all of the benefit go to consumers? - Capturing benefit could incentivize investment - Electricity generation in competitive markets has a fixed cost recovery problem

31 Thank you! Matt Butner University of Colorado Boulder Contact:

32 Proof of firm s incentives Profit Function Π o (S o (p)) = p[s o (p) + θ o W ] C o (S o (p)) (2) First Order Conditions Provide p C (S o (p)) = S o(p) + θ o W d (p) j o S j (p) Comparative Static S o (p) W = θ o Back to picture proof

33 Market Concentration and Diversity Back to MISO details

34 Purchasing Power Agreements Back to MISO details

35 Vertical Arrangements Table: Operations of Utilities with Large Wind Capacity in MISO, 2016 Utility TWh % Wholesale Purchase % Sale for Resale MidAmerican Energy Northern States Power ALLETE, Inc DTE Electric Company Wisconsin Electric Power Basin Electric Power Wisconsin Power & Light Consumers Energy Interstate Power and Light Montana-Dakota Utilities Notes: Capacity is total installed, operating, capacity in megawatts. Wind capacity is the capacity of all wind turbines. All data comes from EIA-860 and EIA-861 for the year TWh stands for terawatt-hour, and represents the thousand of gigawatthours sourced and dispositioned that year. Of the total amount sources, the % Wholesale Purchase represents the amount of electricity they purchased from the wholesale market, the remaining percent (from 100) is the share they generated. The % Sale for Resale is the percentage of total disposition that was sold to a third party (e.g. the wholesale market) the remaining share was sold to retail customers. Back to MISO details

36 Addition control variables - Load - MISO - Net Exports - MISO - Daily maximum temperature - NOAA - Hourly number of binding constraints - MISO - Hourly shadow price of congestion - MISO - Daily Henry Hub natural gas price - YES Energy - Wind forecast error - Yes Energy Back to Specification

37 Three calculations of consumer surplus 1. Perfect Competition, no withholding CS comp = t 1 D t (p) o S ot(p) d t(p) dw t 2. Supply Function Equilibrium, perfect withholding CS SFE = D t (p) 1 ( o V θ ) o t t o S ot(p) d t(p) dw t 3. Observed Withholding Estimates CS obs = t 1 o V D t (p) ˆδ o o S ot(p) d t(p) dw t where ˆδ o is an estimate of So(p) W. Back to Consumer Surplus

38 Summary Statistics of Expected Price Change Table: Analytical Merit Order Effect Mean Std. Dev. Minimum Maximum Observations dp, USD/GWh ,117 dw comp dp, USD/GWh ,117 dw sfe dp comp, USD ,117 dp sfe, USD ,117 dp Notes: come from the theoretical prediction of the impact of 1 GWh of wind on dw the price of electricity with the corresponding assumptions on the price of electricity. comp corresponds to competitive conduct and sfe corresponds to the supply function equilibrium model. The values of dp comp,sfe come from multiplying dp by the GWh of dw wind based electricity. The slopes of supply and demand come from the equilibrium without wind bids and demand less of net exports. The value of o V θo is set equal to the proportion of wind that is generated by diverse market participants in a hour. Back to Consumer Surplus

39 Reconstructing the Equilibrium Back to Consumer Surplus